Fraud metrics

Fraud metrics show key indicators of fraudulent attacks, for example, short click to install times, clicks from suspicious IPs, etc.

Metrics are organized into the following groups:

  • Hardware
  • Click
  • In-app fraud
  • Web fraud (specific to websites)

Based on these groups, we built Combined metrics comes in handy when you first analyze to detect strict, confident and soft fraud.

All indicators are available under the Basic or Premium plan the next day after the user has installed an app or visited your website. See the most accurate data on the third day.

Combined fraud

Combined metrics are convenient at the initial stage of fraud-monitoring, allow to estimate indicators in general and, if necessary, proceed to the analysis of specific metrics.

Combined metrics include the following:

  • Strict Fraud Metric
  • Strict Web Fraud Metric
  • Confident Fraud Metric
  • Soft Fraud Metric
  • Soft Web Fraud Metric
  • Artificial Intelligence Fraud Detector

A detailed description of these metrics is given in the Selectors section

The type of combined metrics (strict, confident, soft) depends on the detected fraud indicators. Strict fraud is devices with precise fraud indicators, for example, Suspicious Devices. While Confident and Soft fraud will get more devices, including those metrics that do not always guarantee 100% frauds.

Combined fraud for mobile apps



Combined fraud for websites



Hardware fraud

Metrics, that allow you to detect fraud from the hardware properties:

  • by knowingly incorrect data about the device or abnormal behavior, for example, when the Android application is installed on the iOS platform;
  • by frequent changes of the launch country or region, by constant reset of key parameters, etc.

Hardware fraud includes the following metrics:

  • Root/Jailbreak devices
  • Suspicious Devices Premium plan
  • Emulated Devices Premium plan

A detailed description of these metrics is given in the Selectors section (Premium metrics description is not available for the Basic plan)

Click fraud

Metrics that indicate fraud in the stage of user interaction with ads and until the app install. Based on the abnormal arrangement of clicks and impressions, we can conclude that we are dealing with a fraudulent device, or that the advertising partner is trying to cheat.

Click metrics are strongly intercorrelated, and abnormal indicators for one of the metrics lead to aggravation of the others.

Click metrics include the following:

  • Stacking Clicks
  • Low CR
  • Quick Installs Premium plan
  • Attribution fraud Premium plan
  • Long Time to Install fraud Premium plan

A detailed description of these metrics is given in the Selectors section (Premium metrics description is not available for the Basic plan)

In-app fraud

Metrics that allow you to recognize a fraud by user behavior after installs, based on abnormal behavior at launches, purchases and other in-app events

In-app fraud metrics include the following:

  • Post-install fraud Premium plan

A detailed description of these metrics is given in the Selectors section ( only for users with premium access to Fraud Scanner)

Web fraud

Metrics that allow you to detect fraud that's typical for websites: active use of proxy, abnormal size of the viewport, or suspicious browsing behavior.

Web fraud metrics include the following:

  • Stacking visits Premium plan
  • Abnormal size Premium plan
  • High PTR (Proxy Through Rate) Premium plan

A detailed description of these metrics is given in the Selectors section ( only for users with premium access to Fraud Scanner)